X.org configuration hell

Laptop: Asus model X540S.

Wireless: Atheros model AR5B125 does seem to work, as well as most other hardware.

Xorg "intel" driver fails because no valid screens are found in FreeBSD.

I have attached files from a working Fedora 25 Linux distribution on this laptop: the output of "lspci -vv" as root and the log file from Xorg, which does work with the "intel" driver in Linux.

EDIT: possibly related https://forums.freebsd.org/threads/59790/


  • Xorg.0.log.txt
    26.2 KB · Views: 393
  • lspci.txt
    23.8 KB · Views: 436
Intel graphics support is way behind Linux's, and probably behind OpenBSD for that matter.

You may have to run CURRENT to get it working, or use vesa graphics which may be painfully slow.

So it looks like you have an E8000

https://forums.freebsd.org/threads/57032/ seemed to have no real solution.

No real answer there. I would try with CURRENT and the latest drm-next. I describe the procedure briefly on a page I have about the yoga2 http://srobb.net/yoga2.html, search for drm-next on the page.
However, I have no idea if it will work with your machine.
Thank you so much for the information!

Intel graphics support is way behind Linux's, and probably behind OpenBSD for that matter.

I am actually much more familiar with OpenBSD than I am with FreeBSD, and I have no problem running it as a guest under virtualization in Fedora Linux, but OpenBSD has no support for hosting virtual machines. (They say it's insecure, and who knows? They may be right.)

When I tried to install OpenBSD 6.0 natively on this laptop, the text mode for the boot kernel messages was overdriving the monitor with an incorrect refresh rate, so I gave up on that.

I have never used NetBSD, so I know nothing in particular about that operating system. Theo DeRaadt was not impressed with their attitude toward security, which is the very reason he forked it and founded OpenBSD with such a strong emphasis on security. I cannot say I am impressed either, going by what they say:

NetBSD Security said:
... The NetBSD source tree contains several millions of lines of code written by many different people and organizations with varying styles and quality. Given the rate of change and the amount of human resources available, it is not possible to manually verify every line of code for correctness. ...
... Several security features are available in NetBSD, ...
... Because high security should not come at the cost of performance and efficiency, not all of these features are enabled by default. ...

In a way this is simply the cold hard truth for any mainstream operating system these days, but the whole attitude is wrong: "This 'is' the way it is and we cannot change."

No, that does not get a pass from me. My attitude is more like this:

#1.) Correctness is a prerequisite for security. Get it right and make sure it stays right.
#2.) Security is not something that is "available." It is something that is not available to one's adversaries.
#3.) Security at the cost of performance and efficiency is a false dichotomy, or rather, a false trade-off. You cannot cut corners in the name of speed and efficiency and claim you are doing it right.